On Tue, Jul 28, 2020 at 08:43:29AM +0200, Thomas Huth wrote:
> On 27/07/2020 12.57, Daniel P. Berrangé wrote:
> > On Fri, Jul 24, 2020 at 05:50:47PM +0100, Peter Maydell wrote:
> >> On Fri, 24 Jul 2020 at 17:46, Philippe Mathieu-Daudé <phi...@redhat.com> 
> >> wrote:
> >>> I guess we were expecting the distrib to update the pkg.
> >>
> >> Apple's view is that you shouldn't be using the sasl header
> >> at all but instead their proprietary crypto library APIs, so
> >> I wouldn't expect them to ever ship something without the
> >> deprecation warnings.
> > 
> > So from pov of our CI, it seems the right answer is to modify the
> > cirrus.yml to install libsasl2 from homebrew:
> > 
> >   https://formulae.brew.sh/formula-linux/libsasl2
> 
> Ok, that one confused me for quite a while, since brew refused to find
> it in the macOS jobs on Cirrus-CI. The solution: This is not a macOS
> package, but a Linux package! Homebrew is apparently also available for
> Linux. There is no libsasl package in homebrew for macOS.

Hah, I totally missed that too.

> So what to do now? I think introducing a libsasl submodule to QEMU just
> for compiling this code on macOS without warnings is also overkill. And
> if I got the answers here right, --disable-sasl is also disliked (since
> this code likely should still be (compile-)tested on macOS).
> So I think I'll go for the same trick as Peter is using for his tests
> and use --extra-cflags='-Werror -Wno-error=deprecated-declarations'.

Seems like despite the deprecation, people just continue to use the
system sasl. I guess the extra-cflags  is reasonable for CI, and it
gives end users a warning that they're relyin on a feature of macOS
that's not considered stable....even if it doesn't appear to have
had any actual changes for 10 years.

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|


Reply via email to